strisunshinewentingwang/pytorch

Pull Requests Commits

Update Transformer.py comments to include a full example (#25411)

Teng Yang committed 5 years ago

e3afe6a4

Updating submodules

svcscm committed 5 years ago

5330b739

Remove forward compat code for serialization format (#25440)

suo committed 5 years ago

0c6ee947

Remove friend dependency on ClassType in InterfaceType (#25617)

suo committed 5 years ago

bb969d5a

Updating submodules

svcscm committed 5 years ago

2eebf084

Enable PiecewiseLinearTransform test on ROCm (#25632)

bddppq committed 5 years ago

b266a079

Kill discover_sparse_tensor_operations. (#25589)

gchanan committed 5 years ago

478440a0

Kill unused enumerate_options_due_to_default.

gchanan committed 5 years ago

9f1a8177

Run clang-format on torch/csrc/distributed (#25647)

pietern committed 5 years ago

5407241b

Fix clang-tidy script (#25652)

pietern committed 5 years ago

ee087a6a

Fix iOS simulator build (#25633)

xta0 committed 5 years ago

14c2492f

Implement initial version of autograd with named tensors (#25604)

zou3519 committed 5 years ago

47cee2dd

Allow TensorMethods.h to include Dispatcher.h (alternative) (#23888)

smessmer committed 5 years ago

79134764

remove protobuf usage from mobile build (#25493)

linbinyu committed 5 years ago

885da48d

switch to rocThrust for thrust/cub APIs (#25620)

iotamudelta committed 5 years ago

4fe85718

Updating submodules

svcscm committed 5 years ago

68b9920c

Name inference rules for relu/relu_/threshold/threshold_ (#25569)

zou3519 committed 5 years ago

0ebbcd95

Fix named tensor printing (#25564)

zou3519 committed 5 years ago

9ea6238b

Add the dynamic quantized LSTM module (#25157)

jianyuh committed 5 years ago

0483d537

Fuse to individual operators to GatherFuse8BitRowwiseQuantFloatMulLengthElim (#25519)

Yinghai Lu committed 5 years ago

4edf77b6

change shape for some ops to reduce variance

hl475 committed 5 years ago

cd4a7cda

Fix binary op name inference to happen before shape checks (#25563)

zou3519 committed 5 years ago

67d64ea9

Name inference rule for torch.cat (#25568)

zou3519 committed 5 years ago

9922e094

Eliminate magic numbers in BatchLinearAlgebra.cu (#25524)

vishwakftw committed 5 years ago

49baeb9d

Don't save `self` in `index` backward (#25594)

zou3519 committed 5 years ago

8edf149f

Name inference for masked_fill_ / masked_fill

zou3519 committed 5 years ago

a6ba4f64

Name inference rule for masked select (#25566)

zou3519 committed 5 years ago

2aef6066

Enable torch.cholesky for batches > 262140 (#24438)

vishwakftw committed 5 years ago

d1e079e2

Get rid of _th_reciprocal_. (#25507)

gchanan committed 5 years ago

0621e2ce

Enable broadcasting of batch dimensions RHS and LHS tensors for lu_solve (#24333)

vishwakftw committed 5 years ago

1e4832ff

Newer Older